Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
This podcast is for information purposes only and should not
be considered professional medical advice.
Speaker 2 (00:05):
Oh if your doctor says stop eating so much and
mac and cheese, I stopped eating mac and cheese.
Speaker 3 (00:13):
Oh.
Speaker 1 (00:13):
I thought you were going to say, I get a
new doctor.
Speaker 3 (00:18):
This is going to destroy medical dramas.
Speaker 1 (00:24):
Yeah, I'd love to have a robot that would fold
my laundry.
Speaker 3 (00:32):
I'm hurry, Condibolu.
Speaker 1 (00:34):
I'm doctor pre Uncle Wally, and.
Speaker 3 (00:35):
This is Health Stuff. On this episode of Health Stuff,
we are keeping it in the Stuff family and talking
to the host of the podcast Science Stuff. Poor Hey Champ.
He's a seasoned podcaster, trained roboticist, and a successful cartoonist.
You're going to enjoy this conversation, so let's get to it.
Speaker 1 (00:55):
Welcome to Health Stuff. I'm so excited to have you
with us today.
Speaker 3 (00:58):
Jorge ours Off colleague.
Speaker 1 (01:01):
Yeah, hello, hello, another member of the Stuffanati.
Speaker 2 (01:05):
That's right, we're all stuff. He's here.
Speaker 1 (01:08):
You know you talk a lot about sign stuff and
all that stuff. But let's take a step back because
I just want to get to know you as a person.
I was really interested in your background that you're actually
from Panama originally.
Speaker 2 (01:21):
Yeah, I was born and raised in Panama.
Speaker 1 (01:22):
Yeah, that to me is so so fascinating. I mean,
I'm curious how coming from Panama did that have a
big impact on the person you are today and inform
your work today.
Speaker 2 (01:35):
Oh? Absolutely? Yeah. Well, first of all, my parents both
worked for the Panama Canal. They're both engineers, computer folks,
and so I think that as an Asian person, that
lucked my destiny and future pretty solidly there. And I
was interested in engineering, you know, I was good at
(01:57):
numbers and things like that, physics, and so yeah, coming
from Panama to here, I first went to study at
Georgia Tech Engineering and then while here I figured out,
oh my goodness, I can study robots for a living.
That's incredible. Sign me up for that. But I think
what beat from Panama. You know, it's just such an
(02:18):
interesting and fun and different culture there. You know, people
are very relaxed, they're very jokey. Humor is a huge
part of the culture. So I think if I had
not been born and raised in Panama, I would be
even less funnier.
Speaker 3 (02:32):
You've had such a fascinating career path, and I think
that speaks to both of us. Just because Prianca is
a doctor who also does comedy and my backgrounds and
immigrant rights organizing and comedy and now podcasting. You went
from engineering to cartoonists to podcast How does that happen?
Speaker 2 (02:52):
I mean, isn't that the career ladder. Everyone has their
life plan there, you know. I'm just going through this
life following what's fun. You know. So I grew up
watching Transformers, the original cartoons, and I was just in
love with robots. So when I saw you can study
that and do that for a living, I was like,
(03:13):
sign me up for that. And then the funny thing
happened was that I was going through graduate school. I
was getting my PhD in robotics, and for that and
just for fun, just as a hobby, I started writing
and drawing comics of the student newspaper there and I
put them online, and at some point I sort of realized, like, oh,
these comics are much more popular than the research papers
(03:36):
I'm putting out, you know, like nobody's calling me about
the research papers or the robots, but there are millions
of people going to this website reading my cartoons. So
I switched to that for a long time, and I
did that for years and years, primarily or in addition
to no, it primarily yeah, wow, it's my full time thing.
(03:59):
You know, com there were books, merchandise, we'ven produced movies
based on the comics and yeah. And then at some
point you're sort of out there and people start calling
you about other projects. Somebody said, hey, do you want
to write a book? I was like, yeah sure. Somebody said, hey,
do you want to pitch us a TV show? I
said yeah, sure, And so that's kind of and that
(04:21):
led to things like the.
Speaker 1 (04:22):
Podcast Lazon you were just doodle, or like did you
grow up drawing? Or like did you have any formal training? Uh?
Speaker 2 (04:31):
Thank you for questioning my credentials. And I feel like
you saw my cartoons and you thought, is this guy?
Speaker 1 (04:40):
No, No, they're really good the video there, they're really great.
Speaker 2 (04:45):
Yeah, thank you. You know what it was when I
was little, Like I said, I grew up in Panama
and my parents worked at the Panama Canal, and so
there were a lot of American families living around the canal,
kind of near where we were. And so one day
my dad was coming home and he saw this garascial,
this American family that was going back to the US,
and they were selling all their stuff and there were
selling these boxes of Archie comics and like old Peanuts
(05:08):
comics from like the sixties, almost maybe even this fifties.
So my dad brought home one day these boxes of
comics and I was like, what is this. I can't, like,
I didn't even know English, I couldn't read English, and
and so I ate those up. And that's partly how
I learned to speak and read English, which explains why
I always say things like oh golly, or he got
(05:34):
talk with a balloon bubbles popping out next to my head.
But so that after that, I was just enamored with
cartoons and drawings since I would just copy them. You know,
there's not a lot to do in Panama, lease in
my house back then, and so I would just draw
and draw and draw. And that carried on in to
my schooling. I would draw in my notebooks and eventually
(05:58):
for this student newspaper there, and so on and so on.
Speaker 1 (06:01):
Yeah, so it sounds like it was just in you.
It was just part of who of who you are.
I can really relate to that. I mean, I started
getting into stand up comedy when I was a resident,
working eighty hours a week, and I just wanted to
blow off some steam.
Speaker 3 (06:18):
You know.
Speaker 1 (06:18):
I just wanted an outlet to do something completely wild
and moving back in time, Like back in high school,
I was voted class clown, so I was always silly,
you know, it was just like in me.
Speaker 2 (06:30):
Yeah, it's funny how art really is kind of like
an outlet for your emotions and what the things you're processing.
I mean, I definitely use with the comics' is therapy,
you know, kind of to figure out what's happening in
my life in that moment. In graduate school, I imagine
it's the same for you in medical school. You're like,
what this is insane? What's going on? Why are people
(06:51):
doing this? Yes?
Speaker 1 (06:53):
Yes, you know it really brings you back to your heart.
And you're formally trained as a I guess would be
a roboticist, so you formally studied robots. And I just
want to switch gears a little bit to talk about that,
because in medicine there's so many incredible things being done
right now with robotics and surgery, and actually this summer
(07:14):
at John Hopkins University, there was a robot that removed
someone's gallbladder completely without human assistance. Yeah, what do you
think about this? As someone who understands robotics.
Speaker 2 (07:28):
Well, it's incredible. You know. I got into robotics just
because I thought robots were cool, you know, and I
was hoping to build the first working Optimus Prime was
really my goal. But well, once I got into it,
I got really interested in actually robusts for health. So
robots have been talked about for helping doctors for a
(07:50):
really long time, helping people like at their home. If
you're bedridden, maybe a robot can help you get around
or help you do things for you.
Speaker 1 (08:01):
Yeah, I'd love to have a robot that would fold
my laundry.
Speaker 2 (08:06):
I don't think that's a health issue, Brianca. That's all right, right, right,
that's right. Yeah, So it's fascinating. Yes, the robots have
been talked about for health for a long time, and
you know, my history with it was robots that could
help doctors in surgery. So this was a really piece
of interesting news for me. But the kind of a
(08:28):
history of it is that for a long time people
use robots mostly to help you in surgery, Like that's
how it started. Like, oh, maybe I can have a
robot hold the camera for me and I can ask
it to point in a different way, And then for
a really long time, starting in the nineties, people figured
out you can use robots to kind of remote control surgery.
(08:49):
So like you still have a surgeon operating kind of
the equivalent of a joystick, but it'd be like a
super complicated joystick, and then maybe in the next room
or in the next continent, you would have these robots
kind of following what the surgeon's doing. So for like
thirty years, that was like the state of the art.
It's like a video game, Yeah, it's basically yeah, like
(09:10):
the doctor would wear a VR goggle and there would
be like stereo cameras in the robot, and so the
idea there is that you know, a doctor could make
kind of crazier moves, like it could move a lot,
and then but the robot would only move like a millimeter,
and so that gives you a lot more precision, a
lot more stability for surgery. So that was basically the
(09:31):
state of the art for a long time, and I
don't think even people were talking about automated surgery, like
that just sounds kind of insane to have a robot
operate on you autonomously.
Speaker 3 (09:42):
It's strange because it feels like the surgeons are almost
training their replacements. Like, is that the goal is the
goal to like have robots do surgery and get rid of.
Speaker 2 (09:53):
Surgeons basically, yeah, I mean that's oh my god, that's
kind of what we're doing right here, or registrating the
robots with our voices right now on how to do
a podcast, right, But yeah, that's that's been the goal
is to get a robot to do it by itself.
And they did it. There's this fascinating experiment and proof
(10:14):
of concept that came out of John Hopkins University by
these engineers and surgeons working together where they had a
robot operate and do a successful surgery on its own.
Like there's nobody like moving a joystick, nobody telling the
robot what to do. It just sort of looked at
the the wound side or what needed to be operated,
(10:37):
and then it just did it.
Speaker 3 (10:39):
Was it on a human or was it on a animal?
Speaker 2 (10:42):
It was on pig, Like it was a pig gallbladder.
So the surgery is that is to remove a gallbladder. Prianca,
you probably know the official name more than I do.
Call a steck to me there you go.
Speaker 1 (10:57):
A systec to me.
Speaker 2 (10:58):
Yeah, So what they did was they trained some robots
to do this surgery. So they the robots kind of
the AIS looked at videos and photos of real surgeons
doing the surgery over and over and over and over again,
and then they could then let the robots lose. Basically
that you said, told the robot, you know, go remove
(11:19):
the gall bladder, and the robots sort of knew what
to do. It figured out, it looked at the scene,
and it made a plan for the surgery. So that's
it's sort of this two step process where they have
an AI that looked at the sort of the photo
of the gall bladder and it figured out, okay, I
need to go first clip this part here, then cut
this part over here, and then put another clip over here.
(11:40):
So it's generated the set of instructions using basically chat
GPT like not not that different than what you're using
to generate means on the internet. And then once they
had that plan, then they had another sort of AI
take those that plan and then plan out what the
robots are we're going to do, and so like you know,
(12:01):
the move the robot this way to cut that part
and then get grab the other tool and then move
over here and put the clip over here. And the
amazing thing is that the robot could do all of
this on its own. And also like sometimes it would
make a mistake, like it would try to grab something
and miss, and the robot wuld be like, oh, I missed,
let me try again. I missed again, let me try again,
(12:21):
or like I didn't cut it the right way, let
me try it again. And so it's this very adaptive
kind of eponymous robot that could do the surgery.
Speaker 3 (12:31):
Wow, I mean, I guess my thought is like, when
this robot's looking at a picture, pictures don't always tell
the full story, right, Are there multiple pictures? Is there
like a thorough way that they can see the actual
thing that they're doing, because like with surgeons, you're looking
you know, you're not just taking a snapshot, You're actually
looking around. You're getting a real big sense of what
(12:53):
the damage is and what has to be done. Or
is it just based on a photo?
Speaker 2 (12:57):
Yeah, you would think there is more to it, but no,
it's just based on the photo and videos. And you know,
they saw the surgeons enough sort of figure out what
to do based on that one that to the picture
that the robots didn't learn like, oh, got a sense
of what was going on sort of like and now
you can take like a just a still picture of
(13:18):
view on the internet and put it through AI and
know generate like a three D model of your head, right,
you know, just like humans when you look at a picture,
you can imagine what that looks like kind of you
turn it around. The same thing is happening in these
AI models.
Speaker 1 (13:32):
Wow, that's wild.
Speaker 2 (13:37):
Yeah. They even tried like changing it up a little bit,
like you know, staining it, like putting a stain on
it so that it's a different color, or putting like
a brand new piece of cloud galbladder, and the robot
would adjust and adapt to it.
Speaker 1 (13:53):
Oh wow, So it like tried to trick the robot
into messing up and it didn't work.
Speaker 2 (13:58):
Yeah, Or like they would started in a weird position,
or they would actually kind of mess it up on purpose,
and the robot would just be like, oh, I think
I know what's going on, and they would adjust and
finish the surgery. One hundred percent success rate that these
robots had, Yeah, we.
Speaker 1 (14:15):
Will see you right after this break.
Speaker 3 (14:21):
So that's going to become commonplace. Like the goal is
for certain surgeries or maybe all surgeries to be done robotically.
Speaker 2 (14:28):
Yeah, that's the idea, And you know, there's a lot
of benefits to that if you think about it, Like,
for example, if you need this very specific surgery, you
don't need the world's best surgeon to fly in from
Zurich and pay bazillions of dollars. You could just have
a robot that's been trained locally to do it on you.
(14:49):
So maybe costs will go down because of it. But
they also think about, you know, places that are hard
to get to that maybe you could just ship a
robot there and then they would do the surgery.
Speaker 1 (15:02):
You know, it's interesting because from my perspective, I see
it from a slightly different lens, Like, isn't it better
to have a surgery done by a robot than let's say,
a resident who's been sleep deprived?
Speaker 2 (15:16):
Yeah, yeah, for sure.
Speaker 1 (15:18):
You know, we hear all the time these cases about
surgeries that were done and then they accidentally left a
sponge in the body and they closed it up, or
you know, they they made a mistake that was completely avoidable.
I'm curious if you know anything about like the error
rate for surgeries like this performed by robots.
Speaker 2 (15:39):
Yes, this is brand new. I think it was relatively
small sample. I think they did it like seventeen times
or something, but one hundred percent success rate. And while
they didn't do it as fast as the surgeons, they
sort of did it more efficiently, meaning like they made
fewer moves that weren't necessary. And they were also a
little bit less jitterate than the surgeons.
Speaker 1 (16:03):
Because they didn't drink all this caff before they went
into there's stone cold surgeons.
Speaker 2 (16:09):
Yeah. Yeah, so they're not faster, but definitely a little
bit more efficient and more stable.
Speaker 1 (16:17):
I mean, it's going to be so interesting how this
plays out. Right, Like again going back to training residents
and they're sleep deprived, they're hungry, but also this is
how they learn, So it's going to be interesting if
like these robots take over, Like how will a general
surgery resident in their first year actually learn how to
(16:37):
do a gall bladder surgery, which is one of the
first surgeries you learn. It's like the bread and butter
of general surgery. I don't know if you have any
thoughts on how it would change the training.
Speaker 2 (16:48):
Yeah, that's a great question. It's sort of like the
question people are posting about the internet in general, like
if everything becomes AI generated, then they're going to be
learning on AI generated things, and people are going to
forget how to do things. It's a fascinating kind of problem.
You know. At some point I imagine maybe the systems
will get good enough where they can also they can
(17:10):
improvise new procedures maybe or find different ways to do it,
in which case then that they can teach themselves and
maybe like we won't need the doctors to learn how
to do it.
Speaker 3 (17:21):
This is going to destroy medical dramas.
Speaker 2 (17:28):
Oh no, we need thirty pints of blood stat and
on no way, the robot already.
Speaker 3 (17:34):
Yeah.
Speaker 2 (17:36):
Well, how does that make you feel, Piranka, I wonder.
Speaker 1 (17:39):
Well, listen, I mean I think we can't stop this.
I see some positive aspects of this, again, like humans
are prone to errors, and I see this as the
answer to that. But I do worry about how the
newer generations are gonna learn these sort of very classic
(18:01):
medicine techniques, and so at some point we're gonna have
to reckon with this. Like I'm imagining a time where
there'll be this rare moment where oh yeah, that doctor
he knows how to do like a traditional gallbladder surgery, like,
let's go find he knows that the old school way,
which everybody knew at one point. So I do think
(18:23):
we can't stop this. I think there's some benefits to it,
but how it's all gonna trickle down, it's gonna be wild.
Speaker 3 (18:31):
Yeah, I mean I have a question about the ethics
of this, like do you have your own reservations.
Speaker 2 (18:35):
Or yeah, you know, I think, like Brianca said, this
is the genies out of the bottle, and so it's
kind of silly to even think about what it'd be
like to go back. One thing that's interesting about these
robot surgeons is that they also program the way for
the surgeon of whoever's there to instruct or interrupt the robots.
So the robot's like doing the surgery, but the doctor
(18:58):
or whoever's there can still say, oh wait, stop, grab
that piece of tendon there and pull it this way.
And if it's train Will, and I think it was
train Will in this case, the robot would actually stop
and do what the person said. I imagine the futures
is some sort of scenario where maybe you have first
year medical students directing a robot on how to do
(19:18):
a surgery and then maybe that's how they learn the
general context. But the actual handling of the scalpel and
the little tweezers maybe will always be done by robot.
Speaker 3 (19:27):
I don't know.
Speaker 1 (19:28):
You know, it's funny. When I was a medical student,
I remember I was on a plastic surgery rotation, or
maybe it was a general surgery rotation. It was some
surgery rotation and I hadn't eaten for probably seven hours,
and I remember the resident or the attending she asked
me to help her suture and my hands because I
(19:51):
hadn't eaten. I was at this point, I'm guessing like
clinically hypogoism it because my hands kept shaking. She was like,
stop shaking your hands, and I remember I was like
I can't, I can't help it. And then she was like, okay,
let me just take over.
Speaker 2 (20:09):
And you're making a great case for the robots.
Speaker 1 (20:11):
Yeah, I mean, I like the idea of that there
will still be clinician oversight of the robots. They won't
be fully autonomous.
Speaker 3 (20:19):
Is that forever though, Pianca? Or is that just in
the initial stages only?
Speaker 1 (20:24):
Who knows.
Speaker 3 (20:25):
I mean, like these waymo, these driverless cars are all over,
you know, like San Francisco, and it's like like there
were there were at some point. I'm sure there were.
They probably tested it with a driver there where the
driver was just in case. But now they're just let's
do it, and that's pretty scary. And but maybe stops
becoming scary when it becomes so commonplace.
Speaker 1 (20:46):
Let's switch sides and look at it from the patient experience, right.
Speaker 3 (20:49):
I was worried, you were about to say from the
robots experience.
Speaker 2 (20:53):
Let's worry about their feelings? Are these were about to
getting enough sleep and food?
Speaker 1 (20:58):
What about their rights.
Speaker 3 (21:00):
They just learned to feel?
Speaker 2 (21:02):
What about their feelings? And who operates in a robot
when they're when they're broken? That's yeah, I want to.
Speaker 1 (21:08):
Get sick too. But I mean, if you flip it,
and let's look at it from the patient experience, right,
Like you're living in a remote village, there's no doctor
for several hours, your only option is being operated on
by a robot, or like you might lose your life.
I mean that at least that's better than not having
(21:32):
any options, I suppose. Yeah, yeah, Like, what what would
you do in that situation if you were the patient?
Speaker 2 (21:38):
You know, oh, I mean, if it's a matter of
life and death, Yeah, I would take the surgery from
a robot.
Speaker 3 (21:44):
Yeah, I understand the idea if it's a remote surgery
and then you have you know, this is your only option.
But if it was like getting a tonsil out tonsilectomy, like,
would you rather a human do it or a robot?
Or like at this stage, are you comfortable enough the
idea of a robot doing it?
Speaker 2 (22:03):
You know, I'm guessing that by the time it becomes
an option, it will have hopefully gone through enough regulation
where it's pretty tried and true and you know, pretty
error free. So probably, by my sense is usually by
the time it becomes an option for me, it's probably
the best option, do you know what I mean?
Speaker 3 (22:22):
Like right right, it becomes so commonplace at that point. Yeah,
I mean, do you see any downsides to this at all?
Speaker 2 (22:29):
Maybe? I mean, robots can always make a mistake. I
mean they're not perfect, especially these sort of like AI
trained you know, they're they're basically going by vibes, you know,
that's kind of what AI is. It's all none of
it is sort of like pre calculated or I mean,
(22:49):
things are structured in a way that they're really good,
but at the end of the day, it's not that
different than a human brain. You know, you put stuff
in and then stuff comes out, and at the time
it's good and so you use it, but there's always
the possibility that it might make a mistake.
Speaker 1 (23:07):
I mean, we're getting closer and closer to this Star
Wars type world. Do you guys remember like the medical
robots and Star Wars, they were exactly like what you
were describing. They were just like fixed all these things.
Speaker 3 (23:23):
It's just so strange of science fiction just becomes how
to or it becomes like a checklist of what has
to be done next and on earth, Like, it's fascinating.
Speaker 2 (23:33):
I think one thing I kind of related to is
getting lazy. So I got Lasig like twenty years ago. Yeah,
and that's a crazy procedure. You grow in they're slicing
your eyeball, they're chuning lasers at it. But it's all
It was all automated, you know, it was all basically
a robot. It didn't have like an arm, but it
was basically a robot. And back then I was like, yes,
(23:54):
sign me up.
Speaker 1 (23:56):
Yeah, I got Lasik too. It was wild. Yeah, and
you can like smell your own burning flash like happening
while it's happening. It's so rad Yeah. Yeah, but it's
crazy because like you can't see, and then by the
end of the surgery you can like see. It's yeah,
it's just so wild. I remember because I was looking
(24:17):
up and you could see the glow orb of the light,
and then by the end of the surgery, I could
see the exact details of the light that they were
shining over me. Like it was so wild.
Speaker 2 (24:28):
Yeah, it's incredible. I love my Laesach. You know, sometimes
I go outside and I look at a tree and
the amount of detail I can see it's overwhelming. I
don't know if this happens to you, but you're like,
oh my gosh, I do think it's seeing a relief
that's oh my god, nature is amazing.
Speaker 1 (24:46):
I totally do that too. I thought I was the
only one like this is so special to here you
say this.
Speaker 2 (24:52):
Yeah, but again, that's kind of an example of of,
you know, a procedure that was kind of automated. It's
very high stakes. It's people's sight, and they figured it
out and they tested a bunch and now it's so commonplace.
It was commonplace twenty years ago to the point where
it almost never made a mistake, and so I imagine the
same thing is going to happen for other surgeries.
Speaker 3 (25:15):
Right, what do you see in the future with robotics,
Like where do you see this going? Like do you
imagine a time where robots are delivering children? Is that
even a possibility of what role do humans have in this?
Speaker 2 (25:28):
Like?
Speaker 3 (25:29):
Are humans gone at this point? Are humans no longer
in the picture in terms of Earth? Is it just
robots at this point?
Speaker 2 (25:35):
Yeah, let's just have robots giving birth to robots and yeah,
no more health. The podcast is over. Yeah, it's going
to it's going to become a mechanical engineering podcast. Yeah,
you know, I think this particular paper, I was surprised
when I read it, and I think it really I
(25:57):
don't I'm not sure, Like the robotic community new is
this close? You know, to have robots fully autonomously do
surgery is pretty incredible, and so I think the genius
kind of out of the bottle, and this is going
to become much more common pla. It's just like LASIK
is now a very commonplace. I think it's going to
be like, oh, you just go into the clinic and
the robot is going to operate on.
Speaker 3 (26:17):
You all surgeries.
Speaker 2 (26:20):
Maybe maybe I think maybe what will still be very
human kind of direct will be the diagnosis, Like you know,
probably the robot is going to diagnose it, but there's
always going to be I think a human person the
doctor going like, yeah, that's a good call, or uh
so I want to get some sleep and some food.
(26:44):
I'll just check this play.
Speaker 3 (26:45):
At that point, the doctor is just an actor playing
the part of it. I can I get know a
wiley to deliver the news, yeah, yeah, I mean, and
also maybe worrying about the health of the patient overall,
I imagine, or being someone that can the patient can
connect to and so they understand what's happening. I imagine
(27:08):
there's always a need for someone to be there and
to kind of make sure it's all going well. But
at that point it's just customer service.
Speaker 1 (27:17):
The question Harry, that you asked about, like, oh, would
a robot ever deliver a baby? I mean, to me,
the process of birth seems so nuanced and so complicated
and so energetic. I can't I can't even fathom that
maybe a robot could do a C section surgery. Since
we seem to be getting surgeries correct so far, but
(27:41):
the idea of like a traditional vaginal delivery assisted by
a robot is almost unfathomable to me. I don't necessarily
think though, it's it's all bad, Like I really think
that there can be some good things that can come
out of this, sure, Like Corey, can you talk to
(28:03):
us about the advancements in artificial limbs, for example.
Speaker 2 (28:08):
Yeah, So, like if you're someone who's had an accident
you lost an arm, or maybe if you're someone who's paraplegic,
had a you know, kind of spinal cord injury and
you lose motion in your limbs, the idea is that
maybe you can use a robotic limb to kind of
replace your missing limb or your ability to control or
even to communicate. Like there are people who can't talk
(28:30):
or can't communicate, so maybe they can use a robot
to communicate for them. And so this has been you know,
sort of going on since the nineteen sixties in the
last century, using robots, using brain signals to control robots,
and so the idea is to help people like that.
And so there's sort of a four levels. So level
zero is just a no prosthetic or just like an
(28:54):
arm or a hand that doesn't move, and you know,
an FPT can still use that to maybe pick something
up or hold something down. Level two is sort of
like a mechanical prosthetic where you maybe have a claw
in it that can open and close in your prosthetic arm,
and that's controlled made by a cable that you attached
through your arm, maybe to your other shoulder, or you
(29:17):
attach it in a way that if you move your
shoulder a certain way, the claw opens and close, sort
of like a puppet almost, And that's actually like a
super popular kind of prosthetic because it's easy, cheap. You know,
you don't have people who have them don't have to
worry too much about the maintenance, or it's not very complicated,
so actually a lot of people sort of opt for that.
(29:38):
So that's level one. Level two is where you're kind
of reading the signals of your peripheral nerves. So like
your brain talks to your spinal cord, and your spinal
cord has nerves that extend out to your hands and
fingers and things like that. But well, if you're missing
an arm, you still have that nerve that kind of
goes up to where you lost the arm, and so
(29:59):
could you maybe read the signals from there and use
that to control like a prosthetic arm that has motors,
that has a claw or has fingers. Actually that's level too,
like can you read those peripheral nerves signals to operate that?
And then level three, which is also very sci fi,
is to actually go directly to the brain, like can
(30:22):
you read signals directly from your brain like the areas
that control the motion of your body, and use that
to control an arm or some sort of mouse on
a computer. So those are sort of the four basic levels.
Speaker 1 (30:36):
Do you see any roll of AIS in artificial whims?
Speaker 2 (30:43):
Yeah, yeah, for sure. So pretty exciting study something that
came out recently from these engineers who do something kind
of different using AI. So what happens is that you know,
the patches on top of your skin are good, but
they're not super accurate. The best thing is to get
(31:04):
the signals directly from the nerve. But the problem with
nerves and generally brain tissue is that if you try
to poke it or like try to put a sensor
on it, they don't like that, and so they use
they'll usually be like ah, you're trying to poke me,
I'm gonna build a lot of scar tissue around it
or get really irritated and then it'll stop working. So
(31:24):
nerve tissue is really sensitive like that. And so what
these scientists figured out recently is that they can graft muscle.
So like if you're missing an arm, you have a
nerve ending that that's not going anywhere. What they do
is they take like a piece of your muscle from
another part of your body, like your thigh muscle, which
is huge, Take a little bit of that muscle, wrap
(31:45):
it around the nerve ending, and now the nerve is happy.
And if that's something to kind of what's called innervate it,
it'll like sort of grow into and make a connection
with those muscle cells, and then the nerve is happy,
and then you can measure the muscle because must not
so picky about you poking it, and so it sort
of amplifies the signal. So it's this crazy procedure where
(32:06):
like they take some muscle graph they wrap it around
the nerve ending, and then they poke. They put electros
in the muscles to read what the nerve is telling
the bit of muscle to do, and then they use
AI to basically train that nerve, train your brain to
control a robot arm. Wow.
Speaker 3 (32:26):
Yeah.
Speaker 2 (32:27):
So they'll like they'll show you like a picture of
a hand opening and closing and be like, okay, can
you try to move your arm even though you don't
have one, try to move it up open and close.
And so you'll use your brain to think like, okay,
I want to move that arm which doesn't exist open
and close, and that will generates some signals. Your brain's
just like oh, wow, I don't know what's going on.
(32:49):
I'm just gonna send the signals down to the nerve
and they'll generate sort of a random set of signals,
and then that will they can train an AI to
be like, okay, this is what for him is trying
to do when he's trying to open and close his arm. Yeah,
and then you might do it for like pointing a finger,
not the middle finger hopefully, but you know, and then
(33:12):
you might be thinking, oh, how can I want to
point that finger? And so then your signals will change
and the AI will learn like, oh, this is what
means when he's trying to point the finger. And so
they've been able to do this, and so you can
basically have an ANAI sort of learn what you're trying
to do right, and then once you once it's trained,
and then you could control an arm and have it
(33:34):
point and open and close.
Speaker 3 (33:36):
I feel like I've said wow, like forty times.
Speaker 2 (33:39):
It's wild. It's super wild.
Speaker 1 (33:42):
So if you were like training this ai about how
you think about making a fist, and it's like monitoring
your neuron patterns to be like okay, when Prianca's brain
goes like this, it makes it fit, like memorize a pattern.
But like if I, while doing that, like thought about
(34:05):
pink elephants, that would get like encapsulated in the training,
so that like whenever I thought about a pink elephant,
my hand would become a vist. Is that like theoretically
what could happen Potentially?
Speaker 2 (34:19):
I mean, that could be a strategy for you to control,
Like if you're finding like it's not working for you,
maybe like the solution could be I mean, I'm not
in it that deeply, but maybe it could be. Like
if you want to close your arm, your prosthetic arm,
maybe the key is to think about a pink elephant,
and so you train it to close every time you
(34:40):
think of it pink elephant, so that if you're out
in the world and you want to grab something, you'd
be like pink elephant and then your hand will close.
Speaker 1 (34:47):
That's wild. Right after this break, we'll be back with more.
We are using AI for now, not just everyday life,
but literally these advancements in medicines, Like it's so pervasive,
(35:08):
you know. And one of the things that I think
about as a physician are the health impacts that these
massive AI data centers have on human beings. Like there
was recently an article in The New York Times about
how data centers are using a lot of water in
(35:29):
order to function and are literally affecting the health of
people in communities out in the middle of America living
close to these centers. I'm curious if you have any
thoughts on that just from a health perspective, and how
like this technology is great and wonderful and I think
it's really important, but at what cost is it coming?
Speaker 2 (35:52):
Yeah? For sure, you know, this is one of those
examples of like oops, right, you know, having consequences that
maybe we didn't expect, you know, or foresee. And what's
happening here is that these ais they don't run cheap
like you know, it's you go on CHATGBT. It's super
(36:13):
easy and fun to like type in something you haven't
answer something back. But what's happening behind the scenes is
that there's millions of people doing the same thing you are,
and so something has to process all this, and so
they have these huge warehouses just full of computers that
do all of these things. Even for the smallest question,
(36:34):
you need to engage the whole sort of neural network.
And so AI has a huge cost to society. They
have these huge warehouses. They suck up an enormous amount
of electricity, and like you said, sometimes there are things
that you don't even think about, like, for example, these
all these computers run really hot, like your phone gets
a little hot if you use it too much. Your
(36:55):
laptop gets too hot. So imagine like a warehouse full
of computers all trying to generate studio Givli memes or
or you know, trying to give you the latest recipe
about something. And so it's enormous I mean I heat.
And the way they deal with that is they have
to cool it, and they use water to cool it.
(37:16):
So it turns out that these giant warehouses used an
enormous amount of water, Like one building can use it
as much water as like thirty thousand homes.
Speaker 3 (37:29):
I mean, it's like an invasive species.
Speaker 2 (37:32):
A little bit. Yeah, Like they try to put it
in places where there aren't a lot of people, but
sometimes those are the places that are most impacted. You know,
in that article, it talks about people who rely on
well water for their water, and so you're basically tapping
into some reservoir under the ground, but then you PLoP
(37:52):
one of these giant centers and it basically sucks up
all the water, leaving no water or maybe contaminating the
water for theeople who need the well water. So yeah, definitely,
I mean not just in things like water, like that's
a very direct impact on health. But if you think
about like if everyone uses these eyes and we're using
all this energy, hopefully they're using renewable energy, but maybe
(38:16):
a lot of them are not, and so you we're
just kind of like accelerating our energy using in the world,
which could lead to an accelerated you know, global warring,
which then will in turn impact people's health.
Speaker 3 (38:30):
How does the water get used to cool down the
like literally used to cool down the serverce like, what
is it doing?
Speaker 2 (38:38):
Yeah, they have internests like with spray bottles, just walking.
Speaker 1 (38:46):
Buckets, throwing them on their back books.
Speaker 2 (38:49):
It's like that video game what is it overcooked? Like,
oh no, the service on fire and then you got
to go grab the bucket and I'm just kidding.
Speaker 1 (38:56):
No.
Speaker 2 (38:56):
It's like a cooling system, you know, like like in
your car you have anti freeze and you have a
cooler system with like tubes that run and they hug
the engine and so you run cool water and that
cools the engine. The same thing happens here. It's like
you have these water just running through pipes, and the
pipes touch the chips that are getting too hot and
so that cools them.
Speaker 3 (39:16):
Okay, yeah, So.
Speaker 1 (39:19):
I mean should we be thinking about this as like
these impacts are just collateral damage to this technology.
Speaker 2 (39:31):
I mean, unfortunately it is collateral damage. But yeah, I
think we just need better planning, you know. And and
I think to think about there's this huge demand for
AI now and everyone's racing to be the person be
the company that has the best AI. Yeah, maybe we
(39:52):
should sort of slow that down a little bit and
think about where all these resources gonna come If we
suddenly double our energy use in the world. Old, how
is that going to impact our health?
Speaker 3 (40:02):
For sure?
Speaker 1 (40:02):
I mean I'm so okay with slowing things down, you know.
I mean, like I remember life pre internet, So I'm
so okay with slowing things down.
Speaker 2 (40:14):
I do feel like it affects our mental health, right,
all this velocity of how things and how we have
this constant pinging that's, you know, interrupting our bods all
the time.
Speaker 1 (40:25):
And I don't think we need it this much. I
mean it's certainly helpful, but I don't I think this
is like really excessive. Yeah, one question I have for you,
so sort of like kind of stepping out now, kind
of wrapping things up a little bit. But I am
so curious about in your study of things that are
happening on a very large collective scale. We know scientifically
(40:48):
and medically, like our nervous systems are connected to each other,
we can actually help each other regulate by picking up
on the frequency of our nervous systems. And so does
this play into your kind of life's philosophy or the
way you kind of look at the world or the
universe as a result of all of your learnings.
Speaker 2 (41:08):
Oh yeah, for sure. I mean, it's definitely humbling to
think about the idea that the universe is fourteen billion
years old, it might be around for trillions of years,
maybe an infinite amount of time. It's humbling to think
about that. Even just our galaxy is one of one
(41:31):
hundred million galaxies that we know about in the universe,
and even our galaxy to get from one side to
the other would take one hundred thousand years going at
the speed of light, the fastest possible speed ever, just
to go from one end to the other. And there's
one hundred million stars and there's probably an earth like
planet in each every single one of those stars on average,
(41:54):
which means there you know, it makes humans both insignificant
and super special. And just this idea that you know,
we're a collection of atoms that can somehow reflect back
on the universe and think about our place in it
and why we're here and wonder how it all works.
(42:16):
To me, it's it's like a miracle. You know, it's
incredible that we can do that, and that that's that's
what we are. We're just little wet ais kind of
trying trying to figure out what's happening and trying to
make ourselves laugh and trying to make ourselves have interesting
thoughts and feel you know, things about the universe, and
(42:41):
it's all sort of like you said, I think it is.
It does point to this idea that you know, our
actions have ripples out there of that it impact other people.
Like we're having this conversation out there's people listening to that.
So our biological machines are creating vibrations in the air
that are getting picked up by this microphone that's turning
(43:03):
that into electricity and into information that's getting zapped across
the world to somebody who's having the reverse process where
the sound is generated and it's impacting their brain and
it's causing ripples on their brain, you know, And that's
incredible to me.
Speaker 1 (43:18):
Yeah, it's a lot of responsibility.
Speaker 2 (43:20):
I guess, well, somebody's got to do it, Prianca might
as well be us. If not us, then who you know.
Speaker 1 (43:28):
To bring it back to health stuff, we're kind of
constantly questioning and discussing, you know, what does it mean
to be healthy? How can we live in a healthy way?
And so I'm curious, how how do you stay healthy?
Speaker 3 (43:44):
Yea, what does health mean to you?
Speaker 2 (43:46):
Great question? One of our first episodes was do I
really have to wait thirty minutes after eating before I
can go swimming? And I had that question for a
long time as a kid, and I talked to several
experts about it, and it turns out you don't really
have to, but you should. And so I think what
(44:08):
that generally tells me about our health is that science
plays a huge role in our health. And you know,
it's not always right, but it's sort of much better
than guessing or making stuff up. And so I think
(44:29):
it's super important right now for people to support science.
You know, it sort of almost feels like science is
under attack. So to me, staying healthy means kind of
kind of paying attention to those experts and supporting science
and ways that I can, and and for me, it
means kind of communicating science, you know, making everyone sort
(44:52):
of be a part of this endeavor to understand our
universe and our world a little more. And fortunately I
have a job where it's sort of my job to
figure out things like my health. Like I just had
an episode about happiness and what are really the factors
that affect your happiness? And so, you know, you can
(45:13):
talk to people about their anecdotes about what make them happy.
But it's a different thing to kind of talk to
somebody who's been running like the longest study about happiness
and history and where he's basically pulled hundreds of thousands
of people and he said, Okay, these are the three
things that are going to make you happy. But yeah,
for me, it means you know, kind of listening and
(45:35):
to the data and strying to stay informed. And you know,
if your doctor tells you get some more exercise, you
get more exercise. If your doctor says stop eating so
much mac and cheese, I stop eating mac and cheese.
Speaker 1 (45:48):
Oh I thought you were going to say, I get
a new doctor.
Speaker 2 (45:54):
I just fire I just fire farre my doctor. That's right.
Speaker 3 (45:57):
Well, he thank you so much for joining us US
and helps. It's nice to have a member of the
family h yes on the show.
Speaker 1 (46:06):
Another stuffhead, yeshead.
Speaker 2 (46:10):
Stuffanati Safinati. I like that one. Let's start the great
chet right now.
Speaker 1 (46:15):
I'm there.
Speaker 3 (46:16):
All right, that's our show for today. A big thank
you to Jorge cham, host of the podcast, signed Stuff,
for joining us, and thank you for listening. We'll be
back next week. I'm Harry Condebolu and I'm doctor Prian Kowally,
and this is health Stuff, a podcast that in a
few years will mostly be about robot doctors operating on cyborgs.
Speaker 1 (46:40):
Health Stuff is a production of iHeart Podcasts. The show
is hosted by me, doctor Prian Kawally, and Hari Kondobol.
Producers are Rebecca Eisenberg, Jenna Cagel, Christina Loranger, Maya Howard,
and Katrina Norvel. Our researcher is Maria Tremarki and our
intern is Katia sobelde Ayala. To send us a question,
(47:01):
you can email us a voice memo at health Stuff
podcast at gmail dot com. Thanks so much for listening.